“The proposed approach involves video taping a scene twice with a hand-held video camera. Each time the recording is performed under different lighting conditions… From the first pass a coarse 3D geometry of the scene is obtained with an associated texture. During the second pass a detailed 3D geometry is obtained but without a texture. Information from these two passes are combined in order to obtain detailed 3D geomtry and texture.”

Modeling Three Dimensional Scenes from Video for Virtual Environments

Marek Czernuskzenko, PhD Thesis, EECS, UIC 1998, p. 47, chapter 3.1